1656
Biochemistry 1982, 21, 1656-1661
strom, B. G., & Vangaard, T., Eds.) pp 221-247, Pergamon Press, Oxford. Bloembergen, N. (1949) Physica (Amsterdam) 15, 386. Blum, H., & Ohnishi, T. ( I 980) Biochim. Biophys. Acta 622, 9-18. Bordas, J . , Bray, R. C., Garner, C. D., Gutteridge, S., & Hasnain, S. S. (1980) Biochem. J. 191, 499-508. Bray, R. C. (1975) Enzymes, 3rd Ed. 12, 299-419. Bray, R . C. (1980) Adv. Enzymol. Relat. Areas Mol. Biol. 51, 107-165. Cammack, R., Barber, M. J., & Bray, R. C. (1976) Biochenz. J . 157, 469-478. Castner, T. G . (1957) Phys. Rev. 115, 1506-1515. Coffman, R. E., & Buetter, G. R. (1979) J. Phys. Chem. 83, 2392-2400. Coughlan, M. P., Betcher-Lange, S. L., & Rajagopalan, K. V. (1979) J. Biol. Chem. 254, 10694-10699. Dutton, P. L. (1971) Biochim. Biophys. Acta 226, 63-80. Good, N . E. (1966) Biochemistry 5, 467-477. Hyde, J. S., & Rao, K. V. S. (1978) J . Magn. Reson. 29, 509-516. Johnson, J. L., Hainline, B. E., & Rajagopalan, K. V. (1980) J . Biol. Chem. 255, 1783-1786. Lowe, D. J., & Hyde, J. S. (1975) Biochim. Biophys. Acta 377, 205-210.
Lowe, D. J., & Bray, R. C. ( I 978) Biochem. J. 169,471-479. Lowe, D. J., Lynden-Bell, R. M., & Bray, R. C. (1972) Biochem. J . 130, 239-249. Lowe, D. J., Barber, M. J., Pawlick, R. T., & Bray, R. C. (1976) Biochem. J . 155, 81-85. Massey, V., & Edmondson, D. (1970) J. Biol. Chem. 245, 6595-6598. Nagler, L. G., & Vartanyan, L. S. (1976) Biochim. Biophys. Acta 427, 78-90. Olson, J. S., Ballou, D. P., Palmer, G., & Massey, V. (1974a) J . Biol. Chem. 249, 4350-4362. Olson, J. S., Ballou, D. P., Palmer, G., & Massey, V. (1974b) J. Biol. Chem. 249, 4363-4382. Palmer, G., & Olson, J. S . (1980) in Molybdenum and Molybdenum-Containing Enzymes (Coughlan, M. p., Ed.) pp 189-220, Pergamon Press, Oxford. Portis, A. M. (1953) Phys. Rev. 91, 1071-1078. Rajagopalan, K. V., Handler, P., Palmer, G., & Beinert, H. (1968) J . Biol. Chem. 243, 3797-3806. Salerno, J. C., & Ohnishi, T. (1980) Biochem. J. 192, 769-78 1. Williams-Smith, D. L., Bray, R. C., Barber, M. J., Tsopanakis, A. S., & Vincent, S. P. (1977) Biochem. J. 167, 593-600.
Determination of the Stoichiometry of Electron Uptake and the Midpoint Reduction Potentials of Milk Xanthine Oxidase at 25 "C by Microcoulometryt Jack T. Spence, Michael J. Barber, and Lewis M. Siegel*
ABSTRACT:
Microcoulometric titrations of several forms of milk xanthine oxidase at 25 OC (bicine buffer, pH 7.7) showed that full reduction of the native enzyme (with a functional Mo center, FAD, and two Fe2S2centers) required six electrons, while reduction of the desulfo enzyme (with FAD, two Fe2S, centers, and a nonfunctional Mo center) required four electrons, and reduction of the desulfo enzyme with its flavin alkylated (with two Fe2S2centers and both nonfunctional FAD and Mo) required two electrons. From these data, the following reduction potentials were determined: Movl/Mov, -375 mV; MoV/MoIV,-405 mV; FAD/FADH2, -280 mV; Fe/S IoxIred,-320 mV; Fe/S IIox,rd, -230 mV. Each of these potentials was shifted significantly in a negative direction (25-75 mV) when native enzyme was titrated coulometrically at pH 8.9 (Ches buffer), in agreement with the pH dependence predicted for these potentials by the results of Barber & Siegel [Barber, M. J., & Siegel, L. M. (1982) Biochemistry (first
paper of three in this issue)]. Pyrophosphate caused positive shifts in the MoV/MoNand Fe/S IIoxId potentials from values found in zwitterionic buffers. The potentials for MoV/Mo'" and Fe/S IIox,rd determined by coulometric titration at 25 OC were 25-30 mV more negative and 25-45 mV more positive, respectively, than the values determined by EPR analysis of enzyme frozen after being poised at defined potentials at 25 OC, whereas the MoV'/MoV, FAD/FADH2, and Fe/S IoxIred potentials obtained by the two methods were in good agreement. It can be concluded that the prosthetic groups of xanthine oxidase possess differing entropies of reduction [dE,/dT = U / ( n F ) ] and , it is therefore not strictly valid to combine results obtained from EPR analysis of some groups in frozen samples with optical spectroscopic measurements of other groups at 25 OC in order to determine the reduction state of the various enzyme centers during titrations and during turnover.
M i l k xanthine oxidase is a dimer of apparently identical M, 140000 subunits, each containing one FAD, one Mo center
(with an associated pterin cofactor; Johnson et al., 1980), and two spectroscopically distinct Fe2S2centers, termed Fe/S I and Fe/S I1 (Bray, 1975). There has been controversy in the literature as to the exact number of electrons which can be taken up by xanthine oxidase. If each Fe/S center takes up one electron and the FAD and Mo centers two each, then xanthine oxidase should take up six electrons per functional half-molecule. Edmondson et al. (1972) and Olson et al. (1974a) observed close to this stoichiometry in reductive ti-
'From the Department of Chemistry and Biochemistry, Utah State University, Logan, Utah 84322 (J.T.S.), and the Department of Biochemistry, Duke University School of Medicine, and the Veterans Administration Hospital, Durham, North Carolina 27705 (M.J.B. and L.M.S.). Received June 25,1981; revised manuscript received December 9, 1981. This work was supported by Grants GM-08347 (to J.S.) and AM-13460 (to L.M.S.)from the National Institutes of Health and Project Grant 7875-01 (to L.M.S.) from the Veterans Administration.
0006-2960/82/0421- 1656$01.25/0
0 1982 American Chemical Society
MICROCOULOMETRY OF XANTHINE OXIDASE
VOL. 21, NO. 7 , 1982
1657
trations with dithionite when the end point was determined by measuring absorbance or EPR changes in the enzyme itself. However, when they followed the disappearance of dithionite, by means of its absorbance at 3 15 nm, these workers found consistent uptake of eight electrons per enzyme half-molecule. There has been speculation as to the nature of the enzyme group responsible for the extra uptake of 2 electrons: reduction of Mo beyond the MoIVstate has been suggested (Coughlan et al., 1980), as has the reduction of a disulfide moiety (Bray, 1975). The interesting possibility must now be raised that the pterin cofactor of the Mo center can undergo oxidation and reduction (Johnson et al., 1980). With the recent introduction of microcoulometry to the investigation of oxidative enzymes, it has seemed appropriate to reinvestigate the problem of the stoichiometry of electron uptake in xanthine oxidase. Microcoulometry utilizes a three electrode potentiostat to poise enzyme samples (containing a mixture of dye mediators to facilitate equilibration between enzyme and electrode) at controlled potentials, with the electron uptake associated with the enzyme at each potential being readily and precisely measured. No chemical reductant, such as dithionite, whose concentration may be difficult to determine due to its facile air oxidation or its ability to undergo side reactions, is required. In addition, microcoulometry can be utilized to measure the reduction potentials of the centers taking up electrons. The measurements are performed at 25 OC, Le., under conditions which can be readily correlated with catalytic function. Such measurements are important for xanthine oxidase, since Olson et al. (1974a,b) have provided strong evidence that internal electron transfer between the various centers of that enzyme is always faster than the rates of reduction and oxidation of the enzyme by substrates. Thus the oxidation states of the various enzyme centers during the catalytic cycle should be determined solely by their relative reduction potentials. Olson et al. (1 974a) calculated such sets of relative potentials, and Cammack et al. (1976) and Barber & Siegel (1982) have determined absolute potentials for the various centers under a variety of experimental conditions. Both groups have utilized EPR spectroscopic analysis of samples frozen in liquid N2 after reduction at 25 OC to follow reduction of most of the enzyme centers. Using such techniques, Barber & Siegel (1982) have reported that the Mo and Fe/S centers of xanthine oxidase, as well as the FAD, accept protons upon reduction, Le., their potentials are pH dependent. They have reached conclusions as to the effects of ligand substitution on the Mo center on proton affinity vs. electron affinity of that center. However, several reports (Williams-Smith et al., 1977; Orii & Morita, 1977) have shown that the process of freezing samples for EPR analysis can result in significant pH changes in the system, with the magnitude of the changes depending upon the nature of the buffer used. Palmer & Olson (1980) have pointed out the possibility of electron reequilibration in frozen samples rendering meaningless the potentials which are to be applied to an enzyme functioning at 25 OC. We have utilized microcoulometry of xanthine oxidase, together with certain of its derivatives which contain prosthetic groups modified so as to render them incapable of accepting electrons under the conditions of our experiments, to determine the reduction potentials of the enzyme centers at 25 OC. We have examined the effects of pH and possible anion binding on these potentials.
al. (1970) described by Barber & Siegel (1982). Concentrations of native and desulfo enzyme were estimated from the A450 by using EmM= 72 mM-' cm-' (Avis et al., 1956). Total active center concentration (functional plus nonfunctional Mo) was taken as twice the enzyme Concentration. The proportion of functional sites in native enzyme was estimated from the ratio (23.5 "C), taking the limiting value of the ratio for fully functional enzyme given by McGartoll et al. (1970). The native enzyme preparation used in this work was 85% functional. Desulfo xanthine oxidase was prepared as described by Massey & Edmondson (1970); the final enzyme preparation was less than 0.5% functional. Desulfo xanthine oxidase containing alkylated FAD was prepared by a modification of the method of McGartoll et al. (1970). Cyanide-treated enzyme, 17 p M in 50 mM Pipes-1 mM EDTA,' pH 6.5, was placed in a Thunberg tube and the solution made anaerobic by repeated evacuation and flushing with Ar. Solid Na2S204 (final concentration 3 mM) and iodoacetamide (final concentration 4 mM) were added from the side arm and dissolved. The closed tube was incubated at 20 "C in the dark for 90 min, and solid cysteine hydrochloride was then added (final concentration 10.5 mM) to quench the reaction. The enzyme was exhaustively dialyzed vs. 50 mM bicine, pH 7.7, and the solution was clarified by centrifugation at 8000 rpm for 30 min at 4 OC. The enzyme was concentrated by ultrafiltration and stored under liquid N2. The degree of flavin alkylation was determined by measurements of the absorbance at 450 and 550 nm, with the assumption that 100%alkylation results in a 37% decrease in the absorbance ratio compared to native or desulfo enzyme (McGartoll et al., 1970). The visible spectrum of the alkylated desulfo enzyme was identical with that previously described for the alkylated native enzyme. Microcoulometry. The basic coulometric technique has been described by Watt (1979) and Watt et al. (1980). A three electrode cell of 3 mL capacity equipped with a Teflon magnetic stirrer, a rubber septum, and a working electrode prepared by electroplating gold onto a double coil of Pt gauze was used. A saturated calomel electrode, separated from the cell solution by a compartment with a Vicor disk, served as reference electrode, and a platinum gauze strip in a compartment separated from the cell solution by a Vicor disk served as counter electrode. A controlled potential was supplied by a PAR 173 potentiostat with an electrometer probe. The enzyme, frozen in liquid N2, was added to a small flask (1 mL capacity) equipped with a rubber septum and allowed to thaw under a stream of nitrogen which had been passed through Oxy-Trap ( < I ppm of 0 2 )The . flask was evacuated and flushed several times with N 2 and then stirred gently at room temperature for 1 h. This procedure reduced the O2 concentration to below the experimental error of the measurements. The cell was filled with 3 mL of solution 0.1 M in KCl at the proper pH and buffer concentration. The solution was deaerated with N 2 and kept under a positive pressure of N,. The mediators were added (5 pL of 30 mM solutions), and the proper potential was maintained with the potentiostat until the current dropped to a small but constant level (1-20 PA). A 5- or 10-pL portion of the enzyme was then transferred from the deaeration flask to the electrochemicalcell with a gas-tight N,-flushed syringe. The time-current curve was recorded and the coulombs required for reduction measured by an electronic
Experimental Procedures Enzyme. Bovine milk xanthine oxidase was isolated from fresh cream by the modification of the procedure of Hart et
I Abbreviations: bicine, N,N-bis(2-hydroxyethyl)glycine;Ches, 2-(Ncyclohexy1amino)ethanesulfonic acid; EDTA, ethylenediaminetetraacetic acid; Pipes, piperazine-N,N'-bis(2-ethanesulfonic acid).
1658
BIOCHEMISTRY
SPENCE,
BARBER,
AND SIEGEL
Table I: Midpoint Reduction Potentials for Xanthine Oxidase Obtained from Microcoulometric Titrations at 25 OC and from EPR Analysis of Frozen Samples
50 mM bicine, pH 7.7 a coulometry couple MoV'/MoV MoV/MoIV FAD/FADH, Fe/S Iox/red
E,d
-375 -405 - 280 -320 -230
EPR c , e Em -373 -371 -283 -310 -255
5 5 mM sodium pyrophosphate, pH 8.2
50 mM Ches, pH 8.9
EPR e Em -453 -427 -340 -350 - 285
coulometry Em -450 -450 -330 -355 -255
EPR Em -355 -355 -294 -343 -303
coulometry Em 9-365 >-335 -315 -345 -260
W S Ilox/red CI All buffers contained 1 mM EDTA. Microcoulometric titrations were performed at 25 "C and midpoint potentials determined as described in the text. Data were obtained for samples titrated potentiometrically in the indicated buffer at 25 "C and then analyzed by EPR spectroscopy of frozen samples. All potentials are in millivolts and are expressed with reference to the standard hydrogen electrode. e Data of Barber & Siegel (1982). Data of Cammack et al. (1976).
integrator in the circuit. Values of the current-time integral were obtained from the current-time curve by the use of a K&E planimeter. Because of small background current level changes, the planimeter method was found to be superior in precision and accuracy to the electronic integrator. After the current reached a small but constant level, the potential was stepped to the next value and the procedure repeated. Each point required a new 5- or 10-pL sample of enzyme. The reference electrode was calibrated after each run against both a standard calomel electrode equipped with a Luggins capillary and a Ag/HCl electrode; it was found to be reproducible with an error of 4 4 mV. The method was checked for both coulometric accuracy and potential with cytochrome c (Sigma, type 111). The error was found to be f 2 5 pC, and Em7 (determined from a Nernst plot) agreed with published values within 10 mV. A mixture of five mediators was used in these experiments: anthraquinone-2-sulfonate (Em7= -225 mV), safranine T (Em7= -289 mV), benzylviologen (E,, = -3 1 1 mV), methylviologen (E,, = -440 mV), and triquat (Em7 = -540 mV). Data Analysis. Midpoint potentials were obtained from the coulometric data as described under Results. Theoretical curves were plotted by using the Nernst equation for n = 1 or n = 2 reduction processes and a given set of potentials by means of a Hewlett-Packard 9825A computer interfaced to a 7225A graphics plotter. Results Native and Modified Xanthine Oxidases (PH7.7). Figure 1 shows the results of microcoulometric titrations of native xanthine oxidase (85% functional) (curve A) and of two modified enzymes: desulfo xanthine oxidase, in which the terminal sulfur ligated to the Mo of functional enzyme has been replaced by an oxygen (Bordas et al., 1980) (curve B), and desulfo xanthine oxidase with its FAD alkylated so as to render the flavin incapable of oxidation and reduction (McGartoll et al., 1970) (curve C). All of the titrations were performed in 50 mM bicine, pH 7.7. Electron addition to the desulfo enzyme with alkylated FAD occurs in a single wave with a midpoint potential of about -270 mV. Reduction is complete at potentials of -370 mV and involves uptake of 2.0 f 0.2 electrons per half-molecule. On the basis of the potentials determined by Barber & Siegel (1982) at pH 7.7, using EPR analysis of frozen samples (see Table I) one would expect the electron uptake in this enzyme derivative in this range of potentials to be associated with reduction of Fe/S I1 and Fe/S I (the alkylated flavin cannot be reduced), There seems to be no electron uptake in the range of potentials expected for reduction of the desulfo Mo centers (average E , for M O ~ ' / M O=' ~-405 mV at pH 7.7; Barber
r---""""""""
I
6 -
u
*
A?-=---
-
#
3 D 0
B
C
0 -103
-zoo
-300 -400 P o t e n t l o 1 (mV)
-500
I I
I
-600
FTGURE 1: Microcoulometrictitration of native xanthine oxidase (curve A), desulfo xanthine oxidase (curve B), and dasulfo xanthine oxidase with alkylated FAD (curve C), in 50 mM bicine-1 mM EDTA, pH 7.7. Each experimental point represents the electron uptake following addition of the following amounts of xanthine oxidase half-molecules to the titration vessel: native enzyme, 2.72 nmol; desulfo enzyme, 2.08 nmol; desulfo enzyme with alkylated FAD, 1.10 nmol. Curves shown were constructed by using the midpoint potentials listed in Table 1. For details see the text.
& Siegel, 1982). This result may indicate an inability of the desulfo Mo to be reduced by the mediator dyes used in the time periods used in these studies; previous experiments have utilized dithionite to reduce desulfo Mo centers, and even with this reductant electron transfer to this center is quite slow (Bray, 1975). The desulfo enzyme, with intact FAD, takes up 4.0 f 0.1 electrons per half-molecule. Reduction occurs in a single wave with a midpoint potential of about -275 mV. These results indicate that the FAD moiety of xanthine oxidase takes up two electrons with an average E , for FAD/FADH2 of about -280 mV (slightly more negative than the average E, values for the two Fe/S centers), a result in good agreement with the potentials determined for the FAD at pH 7.7 by Barber & Siegel (1982) (Table I). Again, there is no electron uptake at potentials more negative than -370 mV, a result which indicates that the desulfo Mo centers are not being reduced. Electron uptake by the native enzyme occurs in two waves: the first corresponds to the four electron reduction process seen with desulfo enzyme, with an indicated midpoint potential of about -275 mV, and the second involves addition of between 1.5 and 2 electrons to each enzyme half-molecule and exhibits an apparent midpoint potential of -390 mV. This potential is similar to (but not identical with) the average functional MoV'/MorVE , at pH 7.7 of -375 mV reported by Barber & Siegel (1982) (Table I). The total uptake of electrons into
VOL. 21, NO. 7 ,
MICROCOULOMETRY OF X A N T H I N E OXIDASE
the native enzyme preparation in two separate experiments (data combined in Figure 1A) was 5.7 f 0.1 electrons per half-molecule, a result which indicates uptake of 2 electrons by each functional Mo center (5.7 - 4.0 = 1.7 electrons per 0.85 functional Mo). The experimental data of Figure 1 were used to find a consistent set of midpoint potentials which would provide the simultaneous best fit to the data for all three enzyme forms, assuming one electron reduction process of Fe/S I and Fe/S I1 in all three enzyme forms, two consecutive one electron reduction processes for FAD in the native and desulfo enzymes, and two consecutive one electron reduction processes for the Mo center in the native enzyme. The data points for the desulfo alkylated enzyme were constrained to fit nc = 1 / [ 1 + ~O(E-EF~/SI)/~~] + 1/[1 + 10("EFe/SlI)/59] The data points for the desulfo enzyme were constrained to fit ne = nc + 1 / [ 1 + 10(E-EFAD,l)/59 + 10(EFAD.2-E)/S9]+ 2/ [ 1 + 1 O(E-EFAD,2)/59+ 1 O(2E-EFAD.I-EFAD,2)/59] The data points for the native enzyme were constrained to fit nA = nB + nc + f/[l + 10(E-E~o.1)/59 + 10(EM0.2-~)/5~] + 2f/ [ 1 + lO(E-Eklo,2)/59+ 10(2"EM,,l-EMo.2)/59] where EF~IS 1, EFe/S 119 EFAD,l, EFAD,2, EMo,lr and EMo,2 are the midpoint potentials for the couples Fe/S I (ox/red), Fe/S I1 (ox/red), FAD (ox/semiquinone), FAD (semiquinone/hydroquinone), Mo(VI/V), and Mo(V/IV), respectively;f = the fraction of xanthine oxidase molecules which contain functional Mo centers (0.85 in Figure 1); n = the number of electrons taken up per half-molecule of xanthine oxidase at solution potential E. All of the data points were used to find a set of midpoint potentials which yielded the minimum total variance between the data points and the respective theoretical curves calculated according to the above equations. The best fit to the data was obtained (using an IBM 1300 series computer with a statistical least-squares program provided by J. P. Chandler, Oklahoma State University) with the potentials indicated in Table I. The computer search indicated that E, for FAD/FADH is much less than that for FAD/FADH2 (it was found that all fits with EFAD,~ - EFAD,, > 30 mV were essentially identical), so that FAD reduction behaves as a two electron process in the coulometric titrations. Thus, only the average E, for the FAD/FADH2 couple could be obtained with any accuracy. (This result is in good agreement with the EPR results of Barber & Siegel (1982), which show that the two potentials for FAD reduction differ by 130 mV at pH 7.7.) The maximum error in the coulometrically determined potentials was f10 mV, as determined by the curve fitting process. Given the errors discussed under Experimental Procedures involved in the microcoulometric measurement itself, it is likely that the potentials of Table I are accurate to at least f20 mV [an error similar to that involved in the potentiometric titrationEPR measurements; see Barber & Siegel (1982)l. It can be seen in Table I that the midpoint potentials obtained at 25 OC from the microcoulometric titration data at pH 7.7 are in good agreement with those obtained by EPR analysis of samples titrated potentiometrically at 25 OC and then frozen for EPR analysis with the following exceptions: the MoV/MorVpotential is shifted more negative by about 30 mV, and the Fe/S II,,,, potential is shifted more positive by
1982
1659
f 4I I
0 -100
-200
-300 Patentla1
-400
-500
-600
(mV)
FIGURE 2: Microcoulometric titration of native xanthine oxidase as a function of pH. Titrations were performed, as described in the text, on enzyme samples in either 50 mM bicine-1 mM EDTA, pH 7.7, or 50 mM Ches-1 mM EDTA, pH 8.9. Each experimental point represents the electron uptake following addition of 2.72 nmol of enzyme half-molecules to the titration vessel. Curves shown were constructed by using the midpoint potentials listed in Table I. For details see the text.
about 25 mV in the frozen samples. Since these species represent the first and last centers to be reduced in the coulometric titrations and electron uptake by them is not greatly masked by uptake into other centers, we can perhaps be more confident with respect to the accuracy of our estimate for these two potentials than with respect to the potentials of the other centers. Effect of p H on Midpoint Potentials. Barber & Siegel (1982), using EPR analysis of frozen samples, have indicated that the midpoint potential for each of the electron carrying centers in xanthine oxidase varies with pH. Williams-Smith et al. (1977) have documented changes in solution pH upon freezing samples in a number of buffers. Although Barber and Siegel used zwitterionic buffers (Good, 1966) to minimize this effect, as suggested by Williams-Smith et al. (1977), it seemed appropriate to see if the same pH effects on midpoint potential are observed if enzyme is both titrated and analyzed at 25 OC. Figure 2 shows the results of a microcoulometric titration of native xanthine oxidase in 50 mM Ches, pH 8.9. The analogous data for native enzyme at pH 7.7 are shown for comparison. It is evident that all of the data points for which equivalent electron uptake occurs are located at more negative potentials at pH 8.9 compared to pH 7.7. The pH 8.9 titration curve, like that at pH 7.7, exhibits two distinct waves of electron uptake: one involving 4 electrons, with an apparent midpoint potential of -320 mV, and the second involving 1.5-2 electrons, with an indicated midpoint potential of -450 mV. Total electron uptake was again 5.7 f 0.1 per enzyme halfmolecule. The potentials determined by EPR analysis of frozen samples are indicated for pH 8.9 and 7.7 in Table I. The EPRderived potentials for the various centers differ at the two pHs as follows (Em,8.9 - Em,7.7): Fe/S 11, -30 mV; Fe/S I, -40 mV; FAD/FADH2, -57 mV; MoV'/MoV,-80 mV; MoV/MoIV,-50 mV. If the pH dependence of the potentials determined by Barber and Siegel is correct, then one should be able to fit the coulometric titration data obtained at pH 8.9 by using the potentials obtained by coulometric titration at pH 7.7 together with the shifts in potential on changing pH estimated by EPR. The curve drawn through the points in the pH 8.2 titration of Figure 2 was calculated by using a set of potentials calcu-
1660
SPENCE, BARBER, AND SIEGEL
BIOCHEMISTRY
I
i'
,
I
1
s
.* . *
-,
J L A
:r:;
i _ i _
- 7! . q 8- -L.
-330 P-.ten-..oi
- AJC
~ . u
-5CO
-GO::
